RuneScape Bench Pits AI Agents Against a Living MMO
RuneScape Bench: MMOs for Agents [video]
A video presentation proposing RuneScape as a benchmark for AI agents: a persistent, open-ended MMO where agents must trade, quest, and cooperate over long horizons rather than solve isolated puzzles. The Hacker News discussion weighs whether a game built for human players makes a fair, measurable test of agent capability.
MMOs for Agents.